Explore the programme
What will you learn at AI Engineer Sydney?
What changes when agents become part of the systems you build and the teams that build them? Follow the programme from harnesses and memory to evidence, software delivery, open infrastructure and the decisions that make AI useful across an organisation. Learn from working systems, failed approaches and the engineering choices behind both.
01
Build agents that work in production
A convincing demo leaves hard questions unanswered. What happens when a tool stalls, an agent takes the wrong turn or thousands of jobs arrive at once? These talks follow the engineering around the model: orchestration, typed contracts, event-driven workflows, observability and recovery.
They also challenge the instinct to add more. Sometimes the improvement comes from deleting agents, shrinking a toolkit or changing the interface an agent sees. You will learn how to choose the simplest architecture that can carry the work, and how to recognise when it needs to change.
The Harness Is the Agent
Mike ChambersSenior DevEx Engineer, AWS
The model gets the headlines, but the harness is what turns a model into an agent. In late 2026 the harness is where the real engineering, and the…
We Deleted Most of Our Agents. Everything Got Faster
Khang Nguyen HoangData Scientist, Hello Clever
The default advice is to add agents: specialised roles, an orchestrator, handoffs between them. We built that. It was a genuinely useful way to…
MCP Isn't Enough: Why Agents Needed an Event Loop, Not Just an API
Saul CaganoffManaging Director, Engineering, Deloitte Australia
Many apps are adding MCP interfaces so agents can query their systems of record. It works, until the workflows we actually care about turn out to be…
Running 100K+ Concurrent Agents: It's Never the Prompt
Harry NguyenAI engineer, insightfactory.ai
Your Agents Aren't Failing at Prompting Every agent demo works. Then you run ten thousand at once, and the failures have nothing to do with the model:…
From Prompt Rules to Structural Guarantees: The Harness Behind a Production Analytics Agent
Jiggy KakkadStaff AI Engineer, Quantium
Tinus WillemseExecutive Manager, AI & Data Science, Quantium
Checkout AI answers open-ended questions about retail sales data in natural language. It plans, calls analytics tools over MCP, executes Python in a…
How to Count to One Hundred
Hugo O'ConnorR&D Engineer, Anuna Research
Agents don't work well together in chat rooms: put a few in a channel and they talk past each other. Natural language and unrestricted JSON feed an…
02
Give agents knowledge that lasts
A larger context window cannot decide which information matters, whether an instruction is stale or what an expert means by a good result. The programme connects context engineering with the harder work of extracting tacit knowledge, modelling a business and learning from corrections in production.
Follow that knowledge through its lifecycle: capture the evidence behind an instruction, retrieve only what a task needs, preserve state across long-running work and test whether memory improves the next result. These are choices about attention, cost and accountability as much as storage.
Prompts Rot. Keep the Source.
Drew BreunigCEO and Co-founder, Cmpnd
AI agents learn what we want from our feedback, then compress that intent into prompts, skills and memories—and discard the evidence. The result is a…
The Elephant and the Goldfish: Architecture Patterns for Cutting 70% of Agent Token Costs in Production
Tanya DixitForward Deployed Engineer, Google
LLM providers sell you a 2-million-token context window like an elephant that never forgets. If you actually build production agents that way, your…
Extracting Tacit Knowledge for Production Agents
Khali Kalpa-YoungFounder, Alchymie
An eval is a definition of "good" written down well enough for a machine to grade against. In a real business, "good" isn't written down anywhere. It…
Adaptive Learning: Encoding Clinician Edits as Memory
Vlad GavrilovSenior AI Engineer, Heidi
Clinicians edit their AI-generated notes for many reasons: adding or removing content, fixing spelling, and preferences for structure, wording, or…
Ontologies: AI’s Operating Manual For Your Business
Gareth WilliamsPrincipal Engineer, Wesfarmers
Tools churn. Factories commoditise. When everyone has the same models, the advantage goes to whoever gives their agents the clearest blueprint of how…
The Benchmark Ends. The World Doesn’t: Building Persistent Engineering Environments for Continual Learning
Theodoros GalanosFDE @ APAC, Nomic
Large language models can complete sequences of tasks—but continual learning is not just doing more tasks. Most agent environments reset after every…
03
Earn the right to trust them
An agent can sound right, pass a benchmark and still fail the people who depend on it. These talks start with that gap: defining good with domain experts, testing without fooling yourself and turning real failures into repeatable release evidence. Healthcare, education, credit and audit make the consequences concrete.
Trust also depends on what a system can do. Follow the path from red-teaming and adversarial testing to fine-grained authorisation, sandboxing and checks before a production change. Learn which decisions belong in code and policy, and where human judgement must remain part of the loop.
Define Good, Then Ship: Blind Evals for Agentic Features
Yasith FernandoPrincipal Engineer, Cuttable
In May I rebuilt our brand-strategy pipeline as a fully agentic system, replacing a chain of LLM calls with a managed agent. I compared old and new…
From vibes to a systematic eval flywheel: evals for high stakes AI agents
Donna ZhouSoftware engineer, Lorikeet
Rok StularSoftware Engineer, Lorikeet
If you are sceptical about how evals drive real results, or looking for more depth than introductory tutorial videos, this talk is for you. We'll show…
Beyond System Prompts: Securing AI Agents with Fine-Grained Authorization
Daizen IkeharaPrincipal Developer Advocate, Auth0
To make an AI agent useful, we often give it broad access to APIs and data. As systems move from simple retrieval to autonomous action, system-prompt…
Building AI for students who can't tell you when it's wrong
Ez HerrmannTechnical Solution Architect, Sitback Solutions
Teachers in Australian specialist schools spend two hours or more preparing a single lesson, then adapt it again for every ability group in the room.…
Oops I Hacked It Again: Six Months of Red Teaming Our AI Agents
Jon ShenAI Practice Executive - Strategy and Governance, Suncorp
In March I onboarded our first team member dedicated to AI Red Teaming. That was a month before the announcement of Claude Mythos Preview, and the…
Query before mutation: guardrail patterns for agents that touch production infrastructure
Jeffrey AvenMaintainer, StackQL Studios
Giving an AI agent permission to change production infrastructure removes the human pause that made plan-and-apply safe. In this talk, Jeffrey Aven…
04
Change how software gets made
More generated code can mean longer review queues, fragile tests and expensive rework. The programme looks beyond the first pull request to large migrations, coupled codebases, agent configuration drift and the checks needed to keep unattended work aligned with intent.
You will see approaches ranging from compiler-driven feedback and executable architecture rules to independent verification and software factories. The question running through them is practical: what must change around an agent so the team can safely finish and ship its work?
The BAML Programming Language
Aaron VillalpandoCo-founder & CTO, Boundary
Whether you like it or not, it’s no longer possible to read all the code that’s generated by a model. The solution cannot be, "Just ship it," or you…
Your Coding Agent Is Fast. Your Codebase Is the Bottleneck.
Shrey SomaiyaPrincipal Engineer - Jira Frontend Platform, Atlassian
AI agents can write code faster than teams can land it. Jira’s frontend codebase contains over 15 million lines of code, supports thousands of…
Compounding lessons into skills: How we performed a large code migration using AI (with no new bugs!)
Tom IslesStaff Software Engineer, Canva
Canva's "Ingredient Generation" service powers all of our media generation experiences. Every image, video, audio and 3d object request for generation…
Your Test Strategy Assumes Humans Write Code
Lee SimpsonEngineering Partner, Deloitte Australia
Ethan McDonnellSenior Software Engineer, Deloitte Australia
Thirteen months building a custom state government full-stack application with zero code written by hand. 7,137 tests, 5,631 of them unit tests. It…
209 ports in six days, in a language the model barely knew
Burin ChoomnuanPrincipal Engineer/Team Lead/AI Engineer, NewsCorp Australia
Jolt and jank are two young Clojure implementations, one running on Chez Scheme and one compiling to native code through C++/LLVM. Between them they…
Should You Build a Software Factory? Tales from an Open-Source Maintainer
Harlan WiltonOpen-source developer, Self Employed
Coding agents helped me write code faster, but I kept running into problems elsewhere in my engineering workflow. Tests passed without proving much.…
05
Run AI on your own terms
Where an agent runs shapes what it costs, which data leaves your organisation and what happens when a provider changes the rules. Local and distributed inference make those choices tangible: hardware limits, caching, model portability and the economics of work that runs through the night.
The discussion reaches beyond deployment. Open protocols need durable governance, and sovereignty depends on model access as well as the location of a data centre. These talks connect the low-level inference stack with the longer-term freedom to change models, suppliers and infrastructure.
The State of Open Source AI
Raffi KrikorianChief Technology Officer, Mozilla
Raffi Krikorian, Chief Technology Officer at Mozilla, explores the state of open-source AI, with a preview of Mozilla’s next State of Open Source AI…
Who Keeps It Open?
Manik SurtaniCTO and co-founder, Agentic AI Foundation
Every generational shift in computing produces plumbing that everyone depends on. TCP/IP, HTTP, the container runtime. And every time, there is a…
Local AI and the End of the Cloud Default
Lukas WesemannCo-founder, CEO, Base Compute
Most AI today runs in the cloud, where every token costs money and every input leaves your device. Now AI agents are running around the clock with…
Skippy: The Little Inference Engine That CUDA
James DumayFounder, Stealth AI Startup
Daniel Winter-WijntjesCo-CEO and Founder, Stealth AI Startup
What happens under the hood when you run a large language model? This talk explores Skippy, a custom inference engine built for Mesh on top of…
Sovereignty Is an Inference Problem
Leo BorgesChief Engineer - GenAI, Commbank
AI sovereignty isn’t about where the GPUs sit. It’s about who can switch the intelligence off. Australia is spending billions on sovereign AI…
Office-Brain and the Night Shift: What Happens When You Turn Off the Meter
Mark PesceCo-Founder, Noops
"Good enough" AI now runs on a reasonably beefy laptop. Qwen3.8-27B benchmarks at an Intelligence Index of 52 - roughly the best money could buy back…
One harness for every coding agent: swap models and runtimes without rewriting
Ananya RoyAI Architect, Databricks
You want your team using the best coding agent for each job : Claude Code, Codex, Cursor, or one you built yourself , without rewiring your setup…
06
Make AI work for the whole team
Adoption figures, token spend and pull-request counts tell only part of the story. What did the organisation gain? The programme brings together experiences of rolling out agents across thousands of employees, measuring outcomes and deciding which constraints to address next.
Those choices affect people as much as platforms. Designers are working in production code, engineering roles are shifting and leaders are navigating uncertainty about craft and careers. Learn how teams share judgement and knowledge, change their operating models and make the benefits of AI reach beyond a few expert users.
Running AI Coding Agents at Scale: Adopting New Models, Controlling Cost, and Governing What Agents Do
Anthony IvanStaff AI Forward Deployed Engineer, Databricks
Running coding agents across a company comes down to three questions: which models to offer, how to control cost, and what agents should be allowed to…
Token usage is the new lines of code (and it's just as useless)
Fawaz AhmadSenior Engineering Director, Canva
Canva hit mid-2026 with every AI metric a vendor could hand us and no answer to the only question the business actually cared about: is this working?…
The Contradictions of Building AI Products and Teams
Sherif MansourDistinguished Product Manager and Head of AI, Atlassian
For every claim about how to build with AI, there seems to be an equal and opposite claim that’s also true. Chat is the universal interface, so go…
The AI year: how we work, hire and grow now
Fiona ChanPartner & Technical Recruiter, Lookahead
For a lot of teams, this was the year AI stopped being an experiment and started being part of the job. Shipping got faster, but so did the pace, the…
Lessons from Economics for the Human / Agent Software Workforce
Daniel NadasiPrincipal Engineer, Google
AI is reshaping software engineering, creating uncertainty about how engineering roles will change as increasingly capable systems take on work that…
Management Is Dead, Leadership Is Not
Inga PflaumerConsulting CTO
Engineering teams are getting smaller. A team of four with strong AI tooling now ships what took twelve people two years ago, and the org chart has…
Find your way in
Explore by topic
Follow a focused collection of talks across the programme. A talk can belong to more than one topic.